Several chemical methods for the sequencing of DNA are known, including the classic method of Maxam and Gilbert. Additionally, the enzymatic "plus-minus" method following Sanger and Coulson, J. Mol. Biol. 94:441-448 (1975), is well known, and is used to elucidate nucleic acid equences of DNA.
An alternative to the "plus-minus" method of Sanger and Coulson, supra, for sequencing of deoxyribonucleic acid (DNA) has been developed by Sanger, Nicklen, and Coulson, (Proc. Nat. Acad. Sci. USA, 74:5463-5467 (1977). This method is based on the use of DNA polymerase-inhibiting nucleoside analogues. Since both arabinonucleotides, and in particular, 2',3'-dideoxynucleotides are used for this purpose, this method is also called the "dideoxy" method.
In the presence of DNA polymerase, 2',3'-dideoxy-nucleoside triphosphates are incorporated into growing oligonucleotide chains at the point of the "correct" nucleoside triphosphates. However, since these do not possess a 3'-hydroxyl group, the chain can not be further elongated after the first triphosphate has been incorporated. The chain growth terminates everywhere a 2',3'-dideoxy-nucleoside triphosphate is incorporated.
This effect is utilized for the sequencing of DNA by dividing a single strand DNA to be investigated into four samples (by, e.g., enzymatic degradation, physical manipulation, and other techniques known to one skilled in the art). Each of these samples is incubated with a short, possibly radioactively-labelled DNA starter molecule in the presence of DNA polymerase, and three different deoxyribonucleoside triphosphates, one of which can be radioactively-labelled, and with a mixture of a fourth nucleoside triphosphate, which will differ in each of the four batches, and the appropriate 2',3'-dideoxynucleotide analogue. After the short start fragment has been hybridized on to the nucleic acid strand, the polymerase begins on the 3'-hydroxyl group of the "start" molecule with the synthesis of the molecule complementary to the DNA sequence to be investigated. The enzyme elongates this molecule until a 2',3'-dideoxy-nucleotide has been incorporated. Thereafter, after termination of the reaction and splitting off of the DNA strand serving as matrix by, e.g., denaturing, a mixture of fragments results which have the same 5' end in all four batches and all display the particular dideoxy analogue used as the 3' end. The gel electrophoretic separation of these molecules gives a band pattern for each batch which reproduces the distribution of the nucleotide corresponding to the particular nucleoside analogue in the newly synthesized DNA. By comparison of the band patterns of the four different batches, which are separated from one another on a gel, one can read off the sequence of the newly synthesized DNA directly. Further, since each nucleotide is complemented by only one other nucleotide, the sequence of the DNA strand serving as matrix can be determined as well.
The advantages of the dideoxy method are, in particular, the simplicity with which it can be carried out and avoidance of an additional incubation and purification step which is necessary in the case of the "plus-minus" method.
Problems arise not only in the case of the sequencing of nucleic acids according to the dideoxy method but also in the case of other sequencing methods which use DNA polymerase, particularly in the case of cytosine-guanine-rich regions. In single strand DNA, cytosine-guanine-rich sequences form stable loops due to internal base pairing. In the case of a gel electrophoretic separation of such fragments, this results in deficient resolution. Furthermore, guanine nucleotides are only moderately stable and, in aqueous solution, tend to form aggregates. This leads to difficulties in the enzymatic polymerization, since insufficient substrate is available on the active points of the enzyme. The result of this is that it is often difficult to sequence cytosine-guanine-rich regions in nucleic acids correctly, when using the dideoxy method.
Therefore, it is an object of the present invention to provide guanosine-nucleotide analogues which are stable, do not enter into self-aggregating units and can be used representatively for 2'-deoxyguanosine-nucleotides as substrate for DNA polymerase, e.g., in the sequencing of nucleic acids according to the dideoxy method.